Echo state networks for online prediction of movement data - Comparing investigations

Abstract

This paper's intention is to adapt Echo State Networks to problems being faced in the field of Human-Robot Interactions. The idea is to predict movement data of persons moving in the local surroundings by understanding it as time series. The prediction is done using a black box model, which means that no further information is used than the past of the trajectory itself. This means the suggested approaches are able to adapt to different situations. For experiments, real movement data as well as synthetical trajectories (sine and Lorenz-attractor) are used. Echo State Networks are compared to other state-of-the-art time series analysis algorithms, such as Local Modeling, Cluster Weighted Modeling, Echo State Networks, and Autoregressive Models. Since mobile robots highly depend on real-time application. © Springer-Verlag Berlin Heidelberg 2008.